基于模型诊断的发展与展望.pdfVIP

第3O卷 第2期 世界科技研究与发展 Vo1．3O No．2 2008年4月 156—160页 WORLD SCI．TECH R＆D Apr．2008 PP．156—160 基于模型诊断的发展与展望 李 冰 姚 锋 钟 炜 黄元亮 一 (1．暨南大学珠海学院计算机系，珠海519070；2．西南交通大学电气工程学院，成都610031； 3．暨南大学电气自动化研究所，珠海519070) 摘 要：在智能故障诊断领域中，基于模型诊断方法已经成为研究者们主要研究的对象。但是，其计算的难解性、诊断结果的数目 的庞大，对实际问题没有统一、寻求适当的模型表示方法是基于模型智能故障诊断必须面临的瓶颈问题。针对国内外专家学者的 研究成就和现状，本文对基于模型诊断方面的主要研究成果进行了简要地介绍，对各种方法的优点和不足进行分析，并对进一步 的研究提出了一些建议。 关键词：最小诊断；核心诊断；静态模型；动态模型；分层诊断；贝叶斯网络 中图分类号：TP31 文献标识码：A Development Prospect of Model—based Diagnosis LI Bing YAO Feng ZHONG Wei HUANG Yuanliang3 (1．Computer Department of Zhuhai College，Jinan University，Zhuhai 519070；2．College of Electrical Engineering，Southwest Jiaotong University，Chengdu 610031；3．Electric Automatization Institute of Jinan University，Zhuhai 519070) Abstract：In the area of intelligent fault diagnoses，the model—based approach is the main approach adopted．However，researchers still have to face the situation that no approach can be found to completely deal with computational complexity and the huge number of solutions to a di— agnostic task of model-based diagnoses as well as no suitable representation within some real—world problems．According to achievements and the present status of research on model—·based diagnoses not only is the relatively mature theory of model·-based diagnosis approaches simply introduced but also their advantages and disadvantages are analyzed，what is more，some pertinent suggestions are proposed for further resear- chcs． Key words：minimal diagnosis；kernel diagnosis；static model；dynamic model；hierarchical diagnosis；bayesian networks 面的分析了基于模型诊断中各种具有代表性